Statistical Prediction Models Research Articles

Construction of battery charge state prediction model for new energy electric vehicles

New energy vehicles have made considerable achievements after nearly a decade of development, but the range and safety of automobile battery packs are still a difficult problem for the industry to overcome. The charge state of the battery can reflect the charging and discharging situation and battery health, so it is significant to construct an accurate prediction model of the battery charge state. In view of this, an RBF neural network is utilized to predict the battery state of charge. Considering that the battery charge state is a nonlinear time series, the study introduces an integrated learning algorithm to further optimize the prediction model. Experimentally, the loss rate of the raised model is almost stable around 150 steps and the loss rate tends to be close to 0. The accuracy of the raised model also reached 81.60%, which is 2.84% and 1.80% higher than the control model, respectively; The recall rate of the raised model reached 77.92%, which was 4.73% and 25.31% higher than the control model, respectively. From the effectiveness test results, the raised model has higher accuracy and faster convergence rate than the control model, and the recall rate of the raised model is also higher, so the raised model has a certain progressiveness.

Comparison of predictive models using logistic regression, classification tree, and structural equation model for severe dengue.

The aim of this study was to compare the predictive performance of three statistical models-logistic regression, classification tree, and structural equation model (SEM)-in predicting severe dengue illness. We adopted modified classification of dengue illness severity based on WHO 1997 guideline. Predictive models were constructed using demographic factors and laboratory indicators on the day of fever occurrence. We developed statistical predictive models using data from two hospital cohorts in Thailand, consisting of 257 Thai children. Different predictive models for each category of severe dengue illness were developed employing logistic regression, classification tree, and SEM. The probability of discrimination of each model for severe output of disease was analyzed with external validation data sets from 55 and 700 patients not used in model development. From external validation using predictors on the day of presentation to the hospital, the area under the receiver operating characteristic curve was between 0.65 and 0.84 for the regression model. It was between 0.73 and 0.85 for SEM models. Classification tree models showed good results of sensitivity, ranging from 0.95 to 0.99. However, they showed poor specificity ranging from 0.10 to 0.44. Our study showed that SEM is comparable to logistic regression or classification tree, which was widely used for more severe form of dengue prediction.

An Investigation in to Sports Analytics and Related Tools

This study explores the application of sports analytics and the utilization of various tools in the sports industry. Sports analytics, also known as sports data analytics, is the application of data analysis and statistical approaches to extract important insights. Additionally, it aids in making well-informed judgments and enhancing performance in the realm of sports. Sports data analytics include a diverse array of methods and approaches, such as statistical analysis, machine learning, data visualization, predictive modeling, and optimization algorithms. The objective of sports business analytics is to get practical insights from a dataset in order to enhance decision-making and foster innovation in the sports industry

P-646 Prognostic statistical model of age-specific anti-Müllerian hormone (AMH) decrease based on hormone values of 23,528 women

Abstract Study question The analysis of anti-Müllerian (AMH) levels, can a useful statistical model be constructed to predict the rate of decrease of AMH levels at different ages? Summary answer A determination of AMH levels can be an objective marker of a ovarian age and a prognostic marker of the rate of ovarian function decline. What is known already The anti-Müllerian hormone is a dimeric glycoprotein produced by granulosa cells of the antral follicles in the ovaries, and AMH participates in the formation of the dominant follicle during folliculogenesis. Serum AMH levels are relatively constant and show little variability between and within the menstrual cycle. The level of serum AMH is proportional to the number of antral follicles and therefore most suitable for determining ovarian reserve. For this reason, AMH is a convenient marker for the biological age of the ovary and for the response to controlled ovarian stimulation (COS). Study design, size, duration A nationwide cross-sectional retrospective study was conducted between 2017 and 2023 and included 23,528 women the ages of 20 - 50, in which AMH levels were measured once. We performed an analysis of the different AMH levels at different ages and constructed a predictive statistical model for the rate of AMH decline. Results were analyzed in subgroups with AMH values < 1 ng/mL, between 1 and 3,5 ng/mL, and > 3,5 ng/mL, at one-year intervals. Participants/materials, setting, methods All women who visited the laboratory for AMH testing were included in the analysis. The samples were collected in 38 points throughout the country and the analysis was carried out in one centralized laboratory using the Ansh Labs (Webster, TX, USA) manual assays, including the ultra-sensitive AMH enzyme-linked immunosorbent assay (ELISA), with a management range of 0,08 - 14,2 ng/ml. The intra-assay variations were 1,97% and 4,00%. The inter-assay variations were 4,63% and 1,98%, respectively. Main results and the role of chance When analyzing all samples in each consecutive year, we found that the rate of AMH decline was approximately linear and inversely proportional to the woman’s age. However, when analyzing the results for the rate of AMH reduction in the different subgroups, we found out that: In women with AMH < 1 ng/ml, the rate of decline with advancing age changes as follows: until age 34, the decline in AMH is very slow, between ages 35 and 45 there is a rapid acceleration of the process.. In women with AMH between 1 and 3,5 ng/ml, the rate of decline with age was approximately constant. In women with AMH > 3,5 ng/ml, the rate of hormone decline is approximately constant but the year-on-year decline rate is decreasing. Based on our analysis and clinical experience, it is our opinion that the rate of AMH decline for the low AMH group has a non-linear characteristic, while ingeneral different groups exhibit different dynamics for the rate of decline. On this basis, we created a statistical model to predict the decrease in AMH levels at different ages. Limitations, reasons for caution Women of only one ethnicity were included in the study. The studied group is heterogeneous, which does not exclude the inclusion of women with a certain pathology and infertility. In some cases, a single determination of AMH could not predict individual differences in rates of hormone reduction. Wider implications of the findings Determination of reference values for AMH at different ages, based on the large volume of examined samples may facilitate the application of AMH in clinical practice. Based on our statistical analysis, creating a prognostic model for decreasing ovarian function and time to menopause appears feasible. Trial registration number NA

Effect of Laser Parameters on Surface Texture of Polyformaldehyde and Parameter Optimization

This research aimed to investigate the influence of laser process parameters on the surface texture of Polyformaldehyde (POM) and to improve its processability and process predictability. A comparative experiment and analysis involving multiple processing parameters, including laser power, scanning speed, and pulse width, were conducted on POM. Statistical prediction models of laser processing POM were established among the laser power, scanning speed, pulse width, texture depth, surface roughness at the bottom of texture, and multi-objective optimization and experimental verification of process parameters were carried out based on the grey-Taguchi analysis method. Experimental results show that the laser power and scanning speed significantly affect the texture depth. Higher laser power and lower scanning speed are conducive to forming depth. The surface roughness at the bottom of the texture increases with the increase in scanning speed and shows a tendency to rise and then fall as the laser power increases. The surface roughness and texture depth obtained under the optimal process parameters(A5B1C1) are 1.373 μm and 466.891 μm, respectively, which were reduced by 10.08 % and increased by 3.42 % compared with the minimum surface roughness and maximum depth in the orthogonal experiments. The validation experiments of the prediction model show that it can meet the reliability requirements, and the errors of the predicted values of depth and surface roughness are 1.86 % and 7.60 %, respectively. The above research provides theoretical and experimental support for the precise control of surface texture prepared by laser processing POM.

An effective parametric sensitivity analysis to improve electrical and thermal energy/exergy efficiency of PVT system using nanofluid

A 3D steady-state model of a photovoltaic/Thermal system with nanofluid is studied to investigate the electrical and thermal performance of PVT by utilizing a combination of statistical modeling and numerical simulation. The 3D CAD (Computer Aided Design) modeling of the photovoltaic thermal (PV/T) is performed using Ansys Space Claim. The current study is based on using a 2-way coupling FSI approach for modeling nanofluid-based PVT systems. Five output responses electrical power, thermal power, electrical exergy, thermal exergy, and entropy generation are chosen against four input parameters, solar radiation (G_sun), ambient temperature (Tamb), mass flow rate (ṁ), and nanoparticles concentration % (φ) to analyze system performance. A statistical prediction model is developed using response surface methodology (RSM) that starts from the design of experiment (DOE). The predicted outcomes obtained from the statistical model and the numerical simulation of the heat transfer model exhibit the existence of good fitness. Furthermore, optimization is performed using RSM for various optimization objectives to suggest the ideal design parameter values for optimal system performance. Sensitivity analysis illustrates that a rise in ambient temperature causes a drop in electrical power production and leads to an increase in thermal power. However, the increase in solar radiation boosts electrical power and causes a drop in thermal exergy, entropy generation increases with an increase in solar irradiation. Similarly, an increase in mass flow rate leads to an increase in thermal power. Multi-objective optimization with the highest composite desirability values depicts the influence of input design parameters and has a favorable effect on the objective of concurrently maximizing electrical and thermal power.

Developing statistical models of aflatoxin risk in peanuts using historical weather data

AbstractAflatoxin contamination in peanuts (Arachis hypogaea L.) is a significant public health risk. Aflatoxin is detected postharvest after inspection of loads associated with grading at peanut buying points, leaving growers and shellers in a precarious position. Stricter limits on aflatoxin contamination could restrict the United States access to international markets. Predicting aflatoxin risk remains challenging, but improved tools could help inform postharvest storage segregation decisions and alert industry stakeholders to seasonal threats. This study aimed to develop and evaluate multiple statistical models that estimate the regional status of peanut aflatoxin contamination based on preharvest weather conditions. Our approach expanded on an existing peanut aflatoxin model for which a new geographic area and time period were tested. Weather variables served as independent variables to predict the risk of aflatoxin as the proportion of samples with greater than 20 ppb and 4 ppb aflatoxin (PGT20 [the proportion of samples with greater than 20 ppb aflatoxin] and PGT4 [the proportion of samples with greater than 4 ppb aflatoxin], respectively) across 10 counties in Georgia for 2018–2022. Best‐performing models were developed through multiple linear stepwise regression explaining more than 72% and 41% of the variability in PGT20 and PGT4, respectively. Model performance further varied whether it was a year of low or high aflatoxin incidence, with temperature observed as a key influencing factor across best‐performing models. This study established an adaptive approach to monitoring and managing aflatoxin risk through statistical predictive modeling, with output targeting farmers, industry, regulators, and public health officials. Future model development will aim to improve interpretation and confidence with in‐season aflatoxin prediction and efficacy testing of this approach across space and time.

Statistical machine learning models for prediction of China's maritime emergency patients in dynamic: ARIMA model, SARIMA model, and dynamic Bayesian network model.

Rescuing individuals at sea is a pressing global public health issue, garnering substantial attention from emergency medicine researchers with a focus on improving prevention and control strategies. This study aims to develop a Dynamic Bayesian Networks (DBN) model utilizing maritime emergency incident data and compare its forecasting accuracy to Auto-regressive Integrated Moving Average (ARIMA) and Seasonal Auto-regressive Integrated Moving Average (SARIMA) models. In this research, we analyzed the count of cases managed by five hospitals in Hainan Province from January 2016 to December 2020 in the context of maritime emergency care. We employed diverse approaches to construct and calibrate ARIMA, SARIMA, and DBN models. These models were subsequently utilized to forecast the number of emergency responders from January 2021 to December 2021. The study indicated that the ARIMA, SARIMA, and DBN models effectively modeled and forecasted Maritime Emergency Medical Service (EMS) patient data, accounting for seasonal variations. The predictive accuracy was evaluated using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R 2) as performance metrics. In this study, the ARIMA, SARIMA, and DBN models reported RMSE of 5.75, 4.43, and 5.45; MAE of 4.13, 2.81, and 3.85; and R 2 values of 0.21, 0.54, and 0.44, respectively. MAE and RMSE assess the level of difference between the actual and predicted values. A smaller value indicates a more accurate model prediction. R 2 can compare the performance of models across different aspects, with a range of values from 0 to 1. A value closer to 1 signifies better model quality. As errors increase, R 2 moves further from the maximum value. The SARIMA model outperformed the others, demonstrating the lowest RMSE and MAE, alongside the highest R 2, during both modeling and forecasting. Analysis of predicted values and fitting plots reveals that, in most instances, SARIMA's predictions closely align with the actual number of rescues. Thus, SARIMA is superior in both fitting and forecasting, followed by the DBN model, with ARIMA showing the least accurate predictions. While the DBN model adeptly captures variable correlations, the SARIMA model excels in forecasting maritime emergency cases. By comparing these models, we glean valuable insights into maritime emergency trends, facilitating the development of effective prevention and control strategies.

Managing Seismic Risk Associated to Development Blasting Using Random Forests Predictive Models Based on Geologic and Structural Rockmass Properties

As mining activities are expected to develop at greater depths, seismic responses to the blasting of development drift segments are expected to increase and present a greater hazard. A database of 379 development blasts was created for a mining site, recording seismic responses related to these blasts and rock mass structural and geologic properties associated with the drift segment. A random forest, multivariate statistical predictive model was developed with 75% of the drift segments. The model's performance was validated by analyzing 100 drift segments that were not used to create the model. The improved understanding of the variation in the intensity of seismic responses to development blasting through the sum of the seismic moment of the events is a clear benefit of random forest model development for the case study. In addition, the development of the predictive random forest model provides a tool for decision-makers to select performance criteria thresholds that they deem acceptable. The threshold selected would depend on the risk appetite of the decision-makers. The proposed approach provides quantitative data on the distribution of seismic hazards associated with development blasting which managers can rely on. Combining the proposed approach with current seismic protocols used at different mine sites could improve our management of seismic risk associated with development blasting. Using the predictive model for the sector and period studied has shown a potential to increase the accuracy, sensitivity, and precision for anticipating a high-intensity seismic response to a development blast.

Generative AI unlocks PET insights: brain amyloid dynamics and quantification.

Studying the spatiotemporal patterns of amyloid accumulation in the brain over time is crucial in understanding Alzheimer's disease (AD). Positron Emission Tomography (PET) imaging plays a pivotal role because it allows for the visualization and quantification of abnormal amyloid beta (Aβ) load in the living brain, providing a powerful tool for tracking disease progression and evaluating the efficacy of anti-amyloid therapies. Generative artificial intelligence (AI) can learn complex data distributions and generate realistic synthetic images. In this study, we demonstrate for the first time the potential of Generative Adversarial Networks (GANs) to build a low-dimensional representation space that effectively describes brain amyloid load and its dynamics. Using a cohort of 1,259 subjects with AV45 PET images from the Alzheimer's Disease Neuroimaging Initiative (ADNI), we develop a 3D GAN model to project images into a latent representation space and generate back synthetic images. Then, we build a progression model on the representation space based on non-parametric ordinary differential equations to study brain amyloid evolution. We found that global SUVR can be accurately predicted with a linear regression model only from the latent representation space (RMSE = 0.08 ± 0.01). We generated synthetic PET trajectories and illustrated predicted Aβ change in four years compared with actual progression. Generative AI can generate rich representations for statistical prediction and progression modeling and simulate evolution in synthetic patients, providing an invaluable tool for understanding AD, assisting in diagnosis, and designing clinical trials. The aim of this study was to illustrate the huge potential that generative AI has in brain amyloid imaging and to encourage its advancement by providing use cases and ideas for future research tracks.

Reliability framework for CLT floors in out-of-plane bending using Monte Carlo simulation

Sustainably sourced engineered wood products (EWP) such as cross-laminated timber (CLT) floor panels are viable alternatives to conventional concrete and steel, however their inherent natural variability increases perceived risks and limits wide-spread adoption. The reliability index (β) is the metric used by design standards to communicate allowable structural risk. Although not included directly in any structural capacity equations, it governs all the partial resistance factors and load amplification factors that inform those design code calculations. CLT floors have significant partial resistance factors due to the natural variability of the material that effectively penalise the calculated design strength of those products. By exploring their reliability there is the potential to allow these products to be designed more efficiently, leading to increased competitiveness among traditional materials and less waste. A novel statistical capacity prediction model that goes beyond a simple averaged-value strength assumption to include consideration of the variability between individual boards within a panel was validated against experimental data. A CLT database of 237 out-of-plane bending test results was collected for the first time to inform an appropriate model error factor. Monte Carlo simulation (MCS) approaches were adopted to predict the reliability levels of CLT panels by defining the material, geometrical and loading parameters as random variables and assigning each a mean, coefficient of variation and appropriate distribution. The target and calculated reliability levels were compared across both Eurocode 5 and the North American codes ANSI/APA PRG 320 and CSA O86 as the preeminent CLT design codes internationally. This study found that the calculated reliability levels met or exceeded both codes in most scenarios. The impacts of width, load ratio, material strength distribution, and CLT layups on the expected failure probabilities were explored with sensitivity analyses. The MCS framework was also applied to the calibration of an appropriate resistance factor for CLT design to inform the potential creation of an Australian CLT design code. A resistance factor of 0.85 was calibrated which falls between the equivalent European and North American code resistance factors.

Accurate Early Detection and EGFR Mutation Status Prediction of Lung Cancer Using Plasma cfDNA Coverage Patterns: A Proof-of-Concept Study.

Lung cancer is a major global health concern with a low survival rate, often due to late-stage diagnosis. Liquid biopsy offers a non-invasive approach to cancer detection and monitoring, utilizing various features of circulating cell-free DNA (cfDNA). In this study, we established two models based on cfDNA coverage patterns at the transcription start sites (TSSs) from 6X whole-genome sequencing: an Early Cancer Screening Model and an EGFR mutation status prediction model. The Early Cancer Screening Model showed encouraging prediction ability, especially for early-stage lung cancer. The EGFR mutation status prediction model exhibited high accuracy in distinguishing between EGFR-positive and wild-type cases. Additionally, cfDNA coverage patterns at TSSs also reflect gene expression patterns at the pathway level in lung cancer patients. These findings demonstrate the potential applications of cfDNA coverage patterns at TSSs in early cancer screening and in cancer subtyping.

High-resolution spatial and spatiotemporal modelling of air pollution using fixed site and mobile monitoring in a Canadian city

The development of high-resolution spatial and spatiotemporal models of air pollutants is essential for exposure science and epidemiological applications. While fixed-site sampling has conventionally provided input data for statistical predictive models, the evolving mobile monitoring method offers improved spatial resolution, ideal for measuring pollutants with high spatial variability such as ultrafine particles (UFP).The Quebec Air Pollution Exposure and Epidemiology (QAPEE) study measured and modelled the spatial and spatiotemporal distributions of understudied pollutants, such as UFPs, black carbon (BC), and brown carbon (BrC), along with fine particulate matter (PM2.5), nitrogen dioxide (NO2), and ozone (O3) in Quebec City, Canada. We conducted a combined fixed-site (NO2 and O3) and mobile monitoring (PM2.5, BC, BrC, and UFPs) campaign over 10-months. Mobile monitoring routes were monitored on a weekly basis between 8am–10am and designed using location/allocation modelling. Seasonal fixed-site sampling campaigns captured continuous 24-h measurements over two-week periods. Generalized Additive Models (GAMs), which combined data on pollution concentrations with spatial, temporal, and spatiotemporal predictor variables were used to model and predict concentration surfaces.Annual models for PM2.5, NO2, O3 as well as seven of the smallest size fractions in the UFP range, had high out of sample predictive accuracy (range r2: 0.54–0.86). Varying spatial patterns were observed across UFP size ranges measured as Particle Number Counts (PNC). The monthly spatiotemporal models for PM2.5 (r2 = 0.49), BC (r2 = 0.27), BrC (r2 = 0.29), and PNC (r2 = 0.49) had moderate or moderate-low out of sample predictive accuracy. We conducted a sensitivity analysis and found that the minimum number of ‘n visits’ (mobile monitoring sessions) required to model annually representative air pollution concentrations was between 24 and 32 visits dependent on the pollutant.This study provides a single source of exposure models for a comprehensive set of air pollutants in Quebec City, Canada. These exposure models will feed into epidemiological research on the health impacts of ambient UFPs and other pollutants.

Predictive approach of health indicators from the physical activity habits of active youth

The aim of this study is to analyse the relationship between sport modalities practiced, physical fitness, body composition, and healthy habits in an active young population, using a statistical model for prediction. A total of 2255 (1528 boys and 727 girls) children and adolescents aged 6–17 years old who were involved in extracurricular sports from rural areas of Spain participated. Physical fitness was assessed through validated field test and, body composition was determinated using bioelectrical impedance analysis. Adherence to the Mediterranean diet was assessed by KIDMED questionnaire. The general sport variable was significant in VO2max when comparing the invasion and combat modalities to the reference level (court/net). The sex and age variables revealed significant differences in all physical fitness and body composition parameters. Health parameters, such as hours of additional practice, adherence to the Mediterranean diet, and previous experience, showed significant differences. The study concludes that the sport modality variables of training, sex, age, and maturational period have an impact on body composition and fitness parameters in this population. Therefore, by focusing on factors associated with lower values in health indicators, we can prevent health problems during adulthood, such as cardiorespiratory deficits.

Surge motion-induced dynamic inflow effects in floating offshore wind turbines: A State Prediction Model

The impact of surge motion on aerodynamic unsteadiness in modern floating offshore wind turbines (FOWT) is a well-recognized phenomenon. When coupled with advanced controllers integrated into these turbines, it can lead to fluctuations in power output. This paper introduces a state-space model that incorporates dynamic inflow effects resulting from surge motion and a PID pitch controller. Additionally, an observation equation is formulated to predict the system power output. Through time series validations, the proposed state-space model demonstrates its ability to accurately capture power responses influenced by surge motions. The developed model serves as a valuable tool for the comprehensive analysis of FOWT, allowing for the exploration of the intricate interplay between unsteady aerodynamics, advanced control mechanisms, and power output.

Short-Term and Imminent Rainfall Prediction Model Based on ConvLSTM and SmaAT-UNet.

Short-term precipitation forecasting methods are mainly divided into statistical forecasting, numerical model-based forecasting, and radar image extrapolation techniques. The two methods based on statistical prediction and numerical model have the disadvantages of being unstable and generating large errors. Therefore, this study proposes the use of deep learning for radar image extrapolation for precipitation forecasting, in particular by developing algorithms for ConvLSTM and SmaAT-UNet. The ConvLSTM model is a fusion of a CNN (Convolutional Neural Network) and LSTM (Long Short-Term Memory network), which solves the challenge of processing spatial sequence data, which is a task that traditional LSTM models cannot accomplish. At the same time, SmaAT-UNet enhances the traditional UNet structure by incorporating the CBAM (Convolutional Block Attention Module) attention mechanism and replacing the standard convolutional layer with depthwise separable convolution. This innovative approach aims to improve the efficiency and accuracy of short-term precipitation forecasting by improving feature extraction and data processing techniques. Evaluation and analysis of experimental data show that both models exhibit good predictive ability, with the SmaAT-UNet model outperforming ConvLSTM in terms of accuracy. The results show that the performance indicators of precipitation prediction, especially detection probability (POD) and the Critical Success index (CSI), show a downward trend with the extension of the prediction time. This trend highlights the inherent challenges of maintaining predictive accuracy over longer periods of time and highlights the superior performance and resilience of the SmaAT-UNet model under these conditions. Compared with the statistical forecasting method and numerical model forecasting method, its accuracy in short-term rainfall forecasting is improved.

Prediction of net primary productivity in the middle-to-high latitudes of Eurasia based on snow and soil temperature

Net primary productivity (NPP) is the net accumulation of organic matter by vegetation through photosynthesis and serves as a key indicator for exploring vegetation responses to climate change. Considering the remote and local impacts of soil heat capacities on vegetation growth through pathways of atmospheric circulation and land–atmosphere interaction, this paper develops a statistical prediction model for NPP from April to June (AMJ) across the middle-to-high latitudes of Eurasia. The model introduces two physically meaningful predictors: the snow water equivalent (SWE) from February to March (FM) over central Europe and the FM local soil temperature (ST). The positive phase of FM SWE triggers anomalous eastward-propagating Rossby waves, leading to an anomalous low-pressure system and cooling in the middle-to-high latitudes of Eurasia. This effect persists into spring through snow feedback to the atmosphere and affects subsequent NPP changes. The ST is closely related to the AMJ temperature and precipitation. With positive ST anomalies, the AMJ temperature and precipitation exhibit an east–west dipole anomaly distribution in this region. The single-factor prediction scheme using ST as the predictor is much better than using SWE as the predictor. Independent validation results from 2009 to 2014 demonstrate that the ST scheme alone has good predictive performance for the spatial distribution and interannual variability of NPP. The predictive skills of the multi-factor prediction schemes can be improved by about 13 % if the ST predictor is included. The findings confirm that local ST is a predictor that must be included for NPP prediction.摘要净初级生产力 (NPP) 是植被通过光合作用积累机物质的净效益, 是探索植被对气候变化响应的关键指标. 考虑到陆面热力异常通过陆气相互作用和大尺度环流对植被生长的影响, 本文研制了欧亚中高纬地区4–6月NPP预测模型. 该模型引入了两个具有明确物理意义的预测因子: 欧洲中部2–3月的雪水当量 (SWE) 和2–3月局地土壤温度 (ST). SWE的正异常会触发异常东传的Rossby波, 导致下游出现位势高度负异常并引起降温. 通过雪和温度的正反馈, 这种异常低温持续到春季并使得NPP下降. ST与随后季节的温度和降水密切相关, 当ST出现正异常时, 该地区的4–6月温度和降水呈现出东西向偶极子异常分布.对比各方案的预测效果发现, ST比SWE有更好的预测能力. 2009–2014年独立后报结果显示, ST方案对NPP的空间分布和年际变率都有很好的预测效果. 交叉检验的结果显示, 多因子方案中引入ST后能提高模型13 %的预测技巧.

Prospective validation of clinical deterioration predictive models prior to intensive care unit transfer among patients admitted to acute care cardiology wards

Objective. Very few predictive models have been externally validated in a prospective cohort following the implementation of an artificial intelligence analytic system. This type of real-world validation is critically important due to the risk of data drift, or changes in data definitions or clinical practices over time, that could impact model performance in contemporaneous real-world cohorts. In this work, we report the model performance of a predictive analytics tool developed before COVID-19 and demonstrate model performance during the COVID-19 pandemic. Approach. The analytic system (CoMETⓇ, Nihon Kohden Digital Health Solutions LLC, Irvine, CA) was implemented in a randomized controlled trial that enrolled 10 422 patient visits in a 1:1 display-on display-off design. The CoMET scores were calculated for all patients but only displayed in the display-on arm. Only the control/display-off group is reported here because the scores could not alter care patterns. Main results. Of the 5184 visits in the display-off arm, 311 experienced clinical deterioration and care escalation, resulting in transfer to the intensive care unit, primarily due to respiratory distress. The model performance of CoMET was assessed based on areas under the receiver operating characteristic curve, which ranged from 0.725 to 0.737. Significance. The models were well-calibrated, and there were dynamic increases in the model scores in the hours preceding the clinical deterioration events. A hypothetical alerting strategy based on a rise in score and duration of the rise would have had good performance, with a positive predictive value more than 10-fold the event rate. We conclude that predictive statistical models developed five years before study initiation had good model performance despite the passage of time and the impact of the COVID-19 pandemic.

A generation-storage coordination dispatch strategy for power system based on causal reinforcement learning

In the backdrop of global energy transformation, power systems integrating high proportions of renewable energy sources are facing unprecedented challenges in operational stability and dispatch efficiency. To address these challenges, this study introduces a generation-storage coordination real-time dispatch strategy based on Causal Power System Dynamic Reinforcement Learning (CPSDRL). Diverging from traditional reinforcement learning approaches, CPSDRL innovatively incorporates causal inference within the state prediction model – the crux of model-based reinforcement learning – thereby establishing the Power Causal Dynamic Model (PCDM). Assisted by the prior knowledge of power systems, the model significantly enhances prediction accuracy and reliability through a two-stage training process. Utilizing PCDM, this study further applies a direct policy search algorithm to optimize the real-time dispatch strategy. Experimental results indicate that the proposed method improves the stability of generation-storage coordination real-time dispatch and exhibits competitive advantages in sample efficiency and computational speed, compared to traditional model-based and model-free reinforcement learning algorithms. This method is expected to enhance the practicality and adaptability of causal reinforcement learning techniques in power system scheduling and control.

Adsorptive removal of perfluorooctanoic acid from aqueous matrices using peanut husk-derived magnetic biochar: Statistical and artificial intelligence approaches, kinetics, isotherm, and thermodynamics

Removal of perfluorooctanoic acid (PFOA) from water matrices is crucial owing to its pervasiveness and adverse ecological and human health effects. This study investigates the adsorptive removal of PFOA using magnetic biochar (MBC) derived from FeCl3-treated peanut husk at different temperatures (300, 600, and 900 °C). Preliminary experiments demonstrated that MBC600 exhibited superior performance, with its characterization confirming the presence of γ-Fe2O3. However, efficient PFOA removal from water matrices depends on determining the optimum combination of inputs in the treatment approaches. Therefore, optimization and predictive modeling of the PFOA adsorption were investigated using the response surface methodology (RSM) and the artificial intelligence (AI) models, respectively. The central composite design (CCD) of RSM was employed as the design matrix. Further, three AI models, viz. artificial neural network (ANN), support vector machine (SVM), and adaptive neuro-fuzzy inference system (ANFIS) were selected to predict PFOA adsorption. The RSM-CCD model applied to optimize three input process parameters, namely, adsorbent dose (100–400 mg/L), pH (3–10), and contact time (20–60 min), showed a statistically significant (p < 0.05) effect on PFOA removal. Maximum PFOA removal of about 98.3% was attained at the optimized conditions: adsorbent dose: 400 mg/L, pH: 3.4, and contact time: 60 min. Non-linear analysis showed PFOA adsorption was best fitted by pseudo-second-order kinetics (R2 = 0.9997). PFOA adsorption followed Freundlich isotherm (R2 = 0.9951) with a maximum adsorption capacity of ∼307 mg/g. Thermodynamics and spectroscopic analyses revealed that PFOA adsorption is a spontaneous, exothermic, and physical phenomenon, with electrostatic interaction, hydrophobic interaction, and hydrogen bonding governing the process. A comparative analysis of the statistical and AI models for PFOA adsorption demonstrated high R2 (>0.99) for RSM-CCD, ANN, and ANFIS. This research demonstrates the applicability of the statistical and AI models for efficient prediction of PFOA adsorption from water matrices using MBC (MBC600).